ABSTRACT Some young adult marijuana (MJ) users report adverse effects of MJ use on cognition that impact daily functioning, with negative consequences such as injury and fatality due to driving while under the influence of MJ. Research on the effects of MJ use on cognition, however, has produced mixed findings. MJ effects on cognition may depend on factors such as history and current severity of marijuana use, time since last MJ use (including possible MJ withdrawal effects), and gender. This R21 aims to address limitations of existing research by (1) starting to develop an algorithm to predict MJ use using smartphone data in regular/heavy MJ users based on ?routine? or ?habitual use?, and (2) examining effects of MJ use on cognition using smartphone- based cognitive testing in the natural environment. Development of an algorithm to predict MJ use would facilitate systematic assessment of MJ effects on cognitive functioning through more efficient scheduling of smartphone cognitive testing among regular/heavy MJ users in relation to daily routines. Cognitive testing by smartphone in the natural environment is an innovative method that has shown validity, and permits sampling of cognitive functioning within and across days in relation to MJ use. This project will enroll non-treatment seeking young adult (ages 18-25) MJ users from the community, representing ?low?, ?regular?, and ?heavy? MJ use, with 50% female at each level of use. Participants will complete a baseline lab assessment, 30-day data collection using smartphone and wearable devices (e.g., wristband), and a debriefing interview. Piloting will optimize the protocol and methods for compliance. Smartphones will collect continuously sensed data (e.g., geolocation) for input to an algorithm to predict MJ use in regular/heavy MJ users. This R21 will identify which types of data, available through smartphone, provide optimal detection of routines in MJ use among regular/heavy users. Smartphone cognitive testing will be administered at various times during acute MJ intoxication and various naturalistically occurring lengths of MJ abstinence to examine effects of MJ use on selected aspects of cognitive functioning in daily life. Development of an algorithm to predict MJ use in regular/heavy MJ users based on smartphone data could, for example, facilitate real-time assessment of MJ effects on cognition through improved sampling of cognition in relation to acute and non-acute effects of MJ use. This R21 will provide the foundation for a research program that aims to examine MJ effects on cognitive functioning in vivo, and could support the development of just-in-time intervention to reduce MJ use. This R21 aligns with NIDA's strategic goal of determining consequences of drug use, and cross-cutting themes of highlighting real-world relevance of research and leveraging mobile health technologies to reduce drug use.